The Agent Skills Directory

Indirect Prompt Injection (HIGH): The skill is designed to read external content (target files) and governance configuration (RULE.md) and follow the 'rules' found within them.
Ingestion points: The skill uses a Read tool to ingest the full content of existing files (Step 2) and RULE.md governance files (Step 3).
Boundary markers: No boundary markers or instructions to ignore embedded commands are specified when processing these files.
Capability inventory: The skill possesses powerful capabilities including file writing (Write), file editing (Edit), and shell-based searching (find, grep).
Sanitization: There is no mention of sanitizing or escaping the content read from these files before the agent processes them as instructions for formatting or governance.
Command Execution (HIGH): The skill explicitly instructs the agent to use shell commands (find . -name "*keyword*" and grep -r "keyword" .) for searching. If the 'keyword' (derived from user input) contains shell metacharacters (e.g., ;, |, &&), it could lead to arbitrary command execution on the host system.
Privilege Escalation / Governance Bypass (MEDIUM): In the 'Error Handling' section for 'RULE.md Forbids Updates', the skill explicitly suggests a workflow to 'override rule'. This provides a direct path for users or potentially malicious files (via indirect injection) to bypass established security and organizational constraints defined in the governance layer.

update-content